Streamlining User Interaction in Tag-Based Conversational Navigation of Knowledge Resource Libraries

摘要

This paper presents an approach for helping users more quickly discover relevant information resources in a tag based system, where each resource is associated with a number of descriptive meta-data tags. Our approach builds an adaptive conversational decision-tree structure to minimize the number of interactive cues required to help a user navigate to resources of interest. Initial experiments demonstrate the potential of the approach, with shallower decision trees supporting better overall interaction performance.